Study On Fault Diagnosis Technology About Chilly Water Set Of Screw

With development of society and improvement of people, s life quality, HVAC systems have been applied in many aspects of production and life. The faults rate increases as the building becomes more intelligent and complex. Fault of HVAC are sure to bring significant influence to life environment and production environment. Understanding the relationship between cause and effect is more difficult than in the past because of complex relationships in the building's processes. So it's necessary to develop the tools that support the operator in decision making for building optimization as well as recovery from a faulty state. This dissertation adopts the combined method including theory analysis and experiment to study the fault detection and diagnosis of chilly water set of screw. Firstly, dissertation adopts comprehensive fuzzy evaluation method to diagnosis fault of chiller and based on Fuzzy Mathematics develops a dedicated software. Secondly, some typical faults were simulated with the intension trying to explain relationships between faults characters and performance parameters. Finally, the dissertation will train the matrix that is characterized by variation of characteristic parameter through artificial neural network method when fault appears. As we know, uncertainty and fuzziness exists between faults of HVAC and causes of them. And the advantage of Fuzzy mathematics is to deal with complex systems, so it is applied in the paper to develop methods for fault detection and diagnosis of water-cooling chiller of screw. Based on Neural Network of human , s brain, man construct Artificial Neural Network that has powerful information-depositing and information-disposing function and that have been used widely to mode-identifying and auto-control. In the third and fourth chapter of this dissertation, we try to apply Artificial Neural Network method to familiar faults diagnosis of HVAC systems. Seven faults such as flux of cooling water of decreasing and as flux of chilly water of decreasing are simulated in this dissertation experiment chapter. We can learn of variation of chiller performance against every fault through experiment data and construct a mathematics model which is characterized by variation of characteristic parameter.